Blog
What we have learned in practice
2026Which AI should I choose? A decision guide for individuals and companiesThe closing part of our guide series. We put the picture together—ChatGPT, Claude, Gemini, Copilot, Perplexity, open models—and give a simple decision order: what to base the choice on, what it costs, and how to roll it out so that your colleagues use it and your data stays safe.2026How to ask well—the quality of the answer depends on the questionMost disappointment with AI comes not from the tool but from the question. Seven simple principles, with plenty of examples, that get a better answer out of any assistant—ChatGPT, Claude, Gemini, Copilot. Not technical tricks, but the way you would explain a task to a new colleague.2026Image, voice, video—the spectacular side of AI at the companyAI does not only write; it draws, speaks and makes films. We look at which tool is for what—images in ChatGPT and Gemini, Midjourney, Adobe Firefly, voice cloning, video generation—what a company can use it for, and where the legal and trust boundaries are.2026Open and European models—when the data cannot leave the buildingNot every AI runs somewhere in America. The world of Mistral, Llama, DeepSeek and downloadable models works by different rules. We look at what an “open” model means, when it is worth running AI on your own computer or server, and what that gives a Hungarian company.2026Perplexity and AI search—when the answer comes with its sourcesSearch has changed. Instead of ten blue links, we get an answer, with citations. We look at what Perplexity is, how it differs from ChatGPT or Google, what it is worth using for in daily work, and where its limits are.2026Copilot—the AI that lives inside Word and OutlookMicrosoft did not ship a new application; it put the assistant into the programs you already use. We look at what Copilot can do in Word, Excel, Outlook and Teams, how it differs from ChatGPT, and what to watch out for when rolling it out.2026Gemini—if your company lives in the Google worldGoogle's assistant is there in Gmail, Docs, Sheets and Calendar, and increasingly in Search as well. We look at what it adds to daily work, what NotebookLM is, and who should choose it instead of ChatGPT or Copilot.2026Claude—the careful colleague who reads it all the way throughAnthropic's assistant is less flashy than ChatGPT, but for long documents, analysis, programming and work connected to company systems, many people choose it. We look at why, and what it is worth using for.2026ChatGPT—the all-rounder most people start withOpenAI's assistant is the best-known AI tool, and not by accident. We go through what it can do in daily work, where it is strong, where caution is advised, and what the difference is between the free, the paid and the business version.2026The AI map—model, application, providerThe first part of our guide series for people who like technology but do not want to become engineers. What the difference is between ChatGPT and the model working behind it, who the major players are, and how to find your way in this fast-changing world.2026MCP—the USB-C of AIAn open standard that finally gives AI access to a company's systems. Eight finance and business processes that are cumbersome at almost every trading company and can now largely be automated—without replacing the existing systems.2026The question isn't whether your employees use AIIt's which tools, and what they paste into them. Free AI tools are in use even at companies that officially have no AI—and that's not a disciplinary issue, it's a leadership one.2026The auditor doesn't ask whether you held a training sessionThey ask whether you can prove it. Hungary's NIS2 law and DORA both require employee security awareness—but little is said about how to demonstrate it. Training and measurement, in plain terms.2026How long can your company last without its systems?The question is not whether there will be an outage, but what it will cost when there is one. Two numbers worth working out in advance—and that nobody at most companies can answer.2026Until now, AI made suggestions. Now it acts.An AI agent doesn't write an answer—it does the task: it signs in to the system, fills things in, sends, makes changes. In plain terms: what this changes in a company's day-to-day work, and where it's worth stopping it.2026You switch on AI in the office. The first surprise won't be what it can do.The AI assistant doesn't break into anything—it finds exactly what the employee already has access to. The problem is that at most companies, nobody knows exactly what that covers.2026triox.hu has been relaunchedOn a new foundation, faster and easier to maintain—briefly, what has changed and why.2026Your boss calls and asks for a transfer. Except it isn't your boss.For most companies, the most tangible effect of artificial intelligence isn't efficiency—it's that fraud has become indistinguishable from the real thing. What you can do about it—mostly without software.2026What Microsoft does not back up for youEmail and files are in the cloud, so they are safe—this is the most common misconception about Microsoft 365. In plain terms: where the provider's responsibility ends, and where yours begins.2026The password that can't be stolenThe major providers are switching on passwordless sign-in one after another. In plain terms: what it is, why it is more secure than even an SMS code, and when it makes sense for a company to make the move.
